No abstract

In this work, we present methods for using human-robot dialog to improve language understanding for a mobile robot agent. The agent parses natural language to underlying semantic meanings and uses robotic sensors to create multi-modal models of perceptual concepts like red and heavy. The agent can be used for showing navigation routes, delivering objects to people, and relocating objects from one location to another. We use dialog clari_cation questions both to understand commands and to generate additional parsing training data. The agent employs opportunistic active learning to select questions about how words relate to objects, improving its understanding of perceptual concepts. We evaluated this agent on Amazon Mechanical Turk. After training on data induced from conversations, the agent reduced the number of dialog questions it asked while receiving higher usability ratings. Additionally, we demonstrated the agent on a robotic platform, where it learned new perceptual concepts on the y while completing a real-world task.

No abstract

Natural language understanding for robotics can require substantial domain-and platform-specific engineering. For example, for mobile robots to pick-and-place objects in an environment to satisfy human commands, we can specify the language humans use to issue such commands, and connect concept words like red can to physical object properties. One way to alleviate this engineering for a new domain is to enable robots in human environments to adapt dynamicallycontinually learning new language constructions and perceptual concepts. In this work, we present an end-to-end pipeline for translating natural language commands to discrete robot actions, and use clarification dialogs to jointly improve language parsing and concept grounding. We train and evaluate this agent in a virtual setting on Amazon Mechanical Turk, and we transfer the learned agent to a physical robot platform to demonstrate it in the real world.

The production of over 800 1.3-GHz superconducting (SC) cavities for the European X-ray Free Electron Laser (EXFEL), the largest in the history of cavity fabrication, has now been successfully completed. In the past, manufacturing of SC resonators was only partly industrialized; the main challenge for the EXFEL production was transferring the high-performance surface treatment to industry. The production was shared by the two companies RI Research Instruments GmbH (RI) and Ettore Zanon S.p.A. (EZ) on the principle of "build to print". DESY provided the high-purity niobium and NbTi for the resonators. Conformity with the European Pressure Equipment Directive (PED) was developed together with the contracted notified body TUEV NORD. New or upgraded infrastructure has been established at both companies. Series production and delivery of fully-equipped cavities ready for cold rf testing was started in December 2012, and finished in December 2015. More than half the cavities delivered to DESY as specified (referred to "as received") fulfilled the EXFEL specification. Further improvement of low-performing cavities was achieved by supplementary surface treatment at DESY or at the companies. The final achieved average gradient exceeded the EXFEL specification by approximately 25%. In the following paper, experience with the 1.3-GHz cavity production for EXFEL is reported and the main lessons learned are discussed.